Problems with the kill web: Moving from C4ISR to SNAI
BlogMarch 27, 2019
Ray Alderman
VITA Standards Organization
WARFARE EVOLUTION BLOG: By now, you know the kill web is a dynamic networked "system of systems," that can act (offensively or defensively) at the speed of computers against our enemy's tactics and strategies on the battlefield. There are a number of technical problems to be solved in communications, computer architectures, sensors, and software, but the engineering brainiacs are working on those. The bigger issues are actually on the operational side of the kill web. Decisions involving many variables must be made in milliseconds or microseconds. The human mind cannot possibly handle all the data from the intelligence sources and sensors, assimilate that data, and make critical decisions in those timeframes. So let?s consider some examples, to expose the complications and contradictions in the kill web.
There are four primary systems in the kill web that target enemy signals in a conflict zone: Signals intelligence including optical imaging (SIGINT and IMINT), cyber warfare (CW), electronic warfare (EW), and finally, all the weapons systems in the fight (missiles, artillery, tanks, planes, ships, UAVs, etc). Each of these systems has a different objective with different reaction times. The SIGINT and IMINT people want to detect, identify, fix, track, and copy the signal for intelligence collection and analysis purposes. The CW people want to hack into the enemy system, suck-out what’s on their hard drive or in memory, manipulate that data, send malware back to that system, and disable it. The EW people want to analyze the enemy signal and effectively jam it. And, the weapons people just want to blow things up. Who decides which system has priority and authority over the target detected by the kill Web?
The basic characteristics of the signal source, speed and proximity, make some decisions easy. If it's an enemy missile or fighter jet, or a column of enemy tanks closing on our soldiers, then the weapons people have priority. The SIGINT/IMINT, CW, and EW folks can play along, but only for as long as it takes our weapon to hit the enemy target. If the target is further away and slower (low threat level), then the weapons and EW people hang back, and the SIGINT/IMINT and CW systems take control over the target.
To read more Warfare Evolution Blogs by Ray Alderman, click here.
You’re probably thinking that it’s even easier to decide which system controls stationary enemy signal targets, but you’d be wrong. If the signals are coming from a ground-based enemy SAM (surface-to-air missile) radar system, and we have friendly aircraft in the vicinity, then the weapons folks get priority over the target again. There are many more examples we could explore here, but you get the idea. These decisions are very complicated.
The time it takes to make these decisions is crucial. We’ve discussed General John Jumper’s concept of time in previous articles. He says that we must find, identify, fix, track, and destroy our enemy’s forces and weapons in 10 minutes or less in a combat area. The kill web is General Jumper's time concept on steroids, operating much faster than humans can think, act, or react. How can we do that? With artificial intelligence algorithms (AI) and high-speed computers. Machines can assimilate the intelligence and sensor data, evaluate that data, and decide which systems have priority and control over the target in fractions of a second. Additionally, machines don’t have historical biases or inter-service rivalries. They won’t question firing a Navy weapon at a ground target threatening Army soldiers. The same goes for the Army firing ground-based missiles at an enemy speedboat threatening a Navy vessel offshore.
However, AI-based decision making won’t happen quickly. We need to get all the computers, sensors, and communications systems in the kill web up and running first, solve the technical problems, and then let the field commanders (Army, Navy, Air Force, Marines) use that data to make the decisions with their brains. That’s called Human-AI collaboration, where the AI system advises the human but the human makes the decision. That might get us down to 10 minutes or a little less. Then, as the AI algorithms become refined and tested, the machines can start making some of the decisions. Eventually, the machines can take over more decision-making processes as younger officers and commanders come into the service, once the older commanders retire and leave. Young people have more faith in technology than older people, since they grew up with it. Older military leaders will retard the acceptance of AI for a few more years.
Once the technology (AI algorithms and high-speed computers) are more mature and trusted, two interesting things will happen. First, the four basic systems in the kill web (SIGINT/IMINT, CW, EW, and weapons) will be integrated together at the hardware and software level. They will no longer be independent systems. Second, the C4ISR concept (command, control, computers, communications, intelligence, surveillance, and reconnaissance) will collapse down to SNAI (sensors, networks, and artificial intelligence). Here’s my theory about how that will happen in a few years…..
The systems will be consolidated and integrated
Sensors (radar, sonar, cameras, signal detectors, infrared detectors, cyber-probes, etc) are just dumb devices today. They simply collect data from their environment and send it over the network to a computer for processing. Then, a human looks at the data and makes decisions or recommendations. Also, sensors can collect more data than the network bandwidth can handle. In the near future, each sensor (SIGINT/IMINT, CW, EW, weapons) will be integrated with a high-speed computer and AI algorithms, and process the data being collected in place. Rather than sending massive amounts of sensor data, only the analysis of that data and the conclusions made by the AI algorithms are sent over the kill web Network. Those results are then compared to what the other sensors in the network have collected and analyzed.
Look at this idea like a military version of Blockchain: If all the sensor/computing systems on the kill web agree, then the decision about which system controls the target is made in fractions of a second. If some nodes on the network disagree, then everyone does their homework again and compare their results until they do agree. Or, the system closest to the target takes control (depending on time constraints and threat level). There are five basic decision algorithms we can use: emergent coordination, the greedy shooter (the system closest to the target takes control), hierarchal coordination, centralized coordination, and consensus coordination. The military operations folks are working on those decision-making algorithms as you read this.
If you look at it logically, EW is just SIGINT in reverse. CW is just network communications and computing in reverse. Jamming an enemy radar system (EW) is just our weapon’s targeting radar in reverse. Those pieces could be integrated together into two basic systems: SIGINT/IMINT/radar/EW and communications/computing/cyber warfare (CC/CW). This hardware will eventually be integrated into one or two boxes, instead of five or six boxes, as semiconductor technology miniaturizes the chips we need.
C4ISR will scale down to SNAI
C4ISR will undergo consolidation and integration too. So, let’s see what’s happening to command, control, computers, communications, intelligence, surveillance, and reconnaissance. The Army is already consolidating command and control into mission control (MC). Computers and communications are slowly being integrated into sensors and networks (SN) through the SOSA program (Sensor Open Systems Architecture) and DARPA’s CODE project (Collaborative Operations in Denied Environments). Intelligence is being replaced by AI. Surveillance has already transitioned to persistent monitoring with the latest Reaper and Global Hawk drones. And reconnaissance is becoming targeting with the latest version of DARPA’s TRACE program (Target Recognition and Adaptation in Contested Environments). C4ISR is transitioning to MCSNAIPMT, but that won’t last long.
When the decision-making algorithms are refined, mission control (MC) will merge into AI. So will persistent monitoring and targeting. Computers and communications are becoming sensors and networks, so that will leave us with SNAI: sensors, networks, and artificial intelligence. You probably doubt my theory, but the U.S. Army just started the Terrestrial Layer System (TLS) project in late February, to integrate SIGINT, IMINT (satellite imagery), CW, and EW systems. Lockheed-Martin just consolidated their SIGINT, CW, and EW groups into a new division: Spectrum Convergence. The Army recently formed the first I2CEWS units (intelligence, information, cyber, electronic warfare, and space). Their mission is hack-jam-sense-shoot (HJSS), and they can do that by integrating their SIGINT/IMINT/CW/EW/weapons systems together. The Air Force has been integrating their SIGINT, EW, and weapons systems in the F-22 and F-35 for years.
As another example, the Army networked the THAAD (Terminal High Altitude Area Defense) radar with the Patriot medium-range air defense radar in 2018. THAAD radar can see farther than Patriot radar, so THAAD can queue the Patriot system to fire a missile at the enemy target even before they can see it on their radar. This is another case of the “greedy shooter” algorithm (the weapon closest to the target takes the shot).
The Navy has been slow to adopt the kill web concept of compressed time, but they are now coming onboard. The chief of naval operations, Admiral John Richardson, recently echoed what Admiral Arleigh Burke said in World War II: “The difference between a good [naval] officer and a poor one is about 10 seconds.” Once a U.S. Navy ship detects an enemy warship in combat, they have about 10 seconds to fix and fire on the target or they are dead. Enemy missiles and high-speed torpedoes don’t take much time to cover 10 or 15 miles (the distance to the horizon).
Before we close this discussion, let’s consider what the kill web is doing to high-level military policy and doctrine. Previous administrations adopted the “principle of proportionality” in war: our troops cannot use more force against an enemy than they can use against us. Proportional response probably cost lives and prolonged the recent conflicts in the Middle East. The objective of this principle was political, not military. The previous administrations didn't want to upset our allies in Europe or the Middle East. But, those days are over. The present administration has adopted the "principle of overwhelming force," and the kill web is the instrument that implements that policy. Future wars will be short, decisive, and extremely lethal for our enemies. If you want to learn more about this element, read Victor Davis Hanson’s book, "Between War and Peace."
So now, you have a general idea about what the kill web concept is doing to C4ISR and military doctrine. For our next expedition into future warfare concepts, I was considering an exhaustive thesis on the latest warfare model coming from the military think tanks: Semiotic Squares (also known as the Greimas Square). This elaborate model is a conceptual tool using Aristotle’s “opposition theory” for structural and relationship analysis, mixed with Harvard Business School’s "matrix analysis," to inject non-linear thinking into tactical and strategic warfare problems. Semiotic Squares is also called "war in a box," and I’d have to create a bunch of box-diagrams to explain it.
In my reading, I have found new information about many topics in previous kill web articles published here. So, let’s do just do an update on platforms and weapons, and review the latest reports on military weapons sales world wide. If I did write my next article on Semiotic Squares, I fear it would only have two interested readers: me and my copy editor.
